I tell new labellers: the first week, your eyes hurt. The second week, you start to see. By the third week you notice things in the street that nobody else notices, and you cannot stop.
Faith Odera
Team lead, Nairobi
Impact Report 2025
Rung by rung
What 2,964 labellers, two centres and 46 clients added up to in 2025, and how many people are standing on a rung.
Letter from the Executive Director
When Joseph Githinji and I started Rungwell, we drew a ladder on a whiteboard in a borrowed office in Kilimani. The bottom rung was a first job. The rungs above it were the things a first job should lead to: a skill, a promotion, a reference, a better job, here or somewhere else. We named the organization after it, and we have measured ourselves against it ever since.
In 2025, 1,031 people stood on the bottom rung for the first time. Forty-one climbed one. At the end of the year 2,964 people worked at Rungwell, and the four-thousandth person since our first cohort joined in October. Our clients accepted 98.7% of what we delivered on the first try, which is the number that keeps the ladder standing: every rung in this report is paid for by careful work that a client chose to buy.
This report also tries to be honest about the rungs that are missing. Most people who join us do not get promoted, because there are far fewer reviewers and team leads than labellers, and there always will be. Too many leave in their first year. We know less than we would like about where people go after us. And labellers doing the hardest work, reviewing distressing content, told us that our wellbeing program is better on paper than on a busy shift. We set out what we are doing about each of these at the end.
Thank you to everyone who sat at a Rungwell workstation this year, in Kilimani, in Cubao and at kitchen tables from Kajiado to Cavite. You are the report.
Nadine Arsenault
Co-founder and Executive Director, Toronto, June 2026
2025 at a glance
Who we hire
We recruit through community youth programs, colleges and parishes in the neighbourhoods around our centres, and since 2020 through Rungwell at Home in towns an hour or more away. We look for people who have finished secondary school, read carefully and want to learn; we do not ask for a degree or for experience. Three in four people who joined in 2025 had no formal job in the year before they applied.5
| Nairobi | Manila | Total | |
|---|---|---|---|
| Working on site | 1,186 | 838 | 2,024 |
| Working through Rungwell at Home | 556 | 384 | 940 |
| All labellers | 1,742 | 1,222 | 2,964 |
| Women | 51.9% | 56.6% | 53.8% |
| Median age on joining | 23 | 24 | 23 |
| No formal work before joining (joined 2025) | 81% | 74% | 78% |
| Living with a disability | 2.4% | 3.1% | 2.7% |
| Joined in 2025 | 612 | 419 | 1,031 |
| Left in 2025 | 498 | 364 | 862 |
Pay
Everyone at Rungwell is an employee, with a contract, a payslip, health cover from the first day, paid leave and statutory contributions. Nobody is paid as a contractor, and no work is passed to a subcontractor. Labellers are paid monthly, with a base for every scheduled shift, and we pay for the bus home after late shifts, a lunch allowance on site, and the laptop and data for Rungwell at Home.
We report earnings as a multiple of the legal minimum wage that applies to the work, because the currencies, costs of living and minimum wages of Nairobi and Metro Manila are too different for a single figure to mean anything. A labeller in Nairobi who worked a full month in 2025 earned a median of 2.3× the minimum; in Manila, 1.6×.2
| Nairobi, labeller | 2.3× |
|---|---|
| Nairobi, senior labeller | 2.9× |
| Nairobi, quality reviewer | 3.6× |
| Manila, labeller | 1.6× |
| Manila, senior labeller | 1.9× |
| Manila, quality reviewer | 2.4× |
The ladder
The ladder is narrow at the top. For every team lead there are about seventy labellers, and that will not change, because a team lead’s job is to look after a team. What we can do is make the rungs real: a published path from labeller to senior labeller to reviewer to trainer or team lead, assessed on quality and on nothing else, with training paid for in work time.
| Year | Promotions |
|---|---|
| 2016 | 2 |
| 2017 | 9 |
| 2018 | 17 |
| 2019 | 24 |
| 2020 | 26 |
| 2021 | 39 |
| 2022 | 44 |
| 2023 | 48 |
| 2024 | 37 |
| 2025 | 41 |
| Since 2016 | 287 |
Promotions fell in 2024 because headcount grew more slowly and few senior roles fell vacant. They recovered a little in 2025, mostly in Manila, where the new ninth floor needed reviewers.11
Faith joined the first Nairobi cohort in March 2016. She became a senior labeller in 2017, a quality reviewer in 2018 and a team lead in 2021, and leads quality on confidential projects in the centre’s restricted rooms. The object she chose is the red pen she marks calibration sheets with.
After Rungwell
A ladder that only leads to more of the same is not much of a ladder, so we ask people who leave where they went. Of those who left in 2024 and answered our survey a year later, 71% were working or studying.7
| Working in digital or office jobs | 34% |
|---|---|
| Working in other jobs | 22% |
| Studying | 15% |
| Looking for work | 21% |
| Caring for family, or other | 8% |
Wellbeing
Some of our work means looking at things nobody should have to look at, so that fewer people have to. Sensitive work is graded in three tiers, from S1 (occasional distressing material) to S3 (sustained exposure, restricted rooms only), and every tier comes with wellbeing time in the shift, access to counsellors on site and by phone, and the right to step away from a project without losing pay for the shift.
- Counselling sessions
- 9,412
- with 21 counsellors, 12 in Nairobi and 9 in Manila
- Labellers who said the program is good
- 71%
- annual survey, 2,207 replies
- Who said it is easy to use on a busy shift
- 44%
- same survey
The last two numbers do not sit comfortably together, and they are why we have committed to an independent review of the program.
The work
We delivered work to 46 clients in 2025, from agricultural imaging companies and retailers to language-model developers, platforms and makers of medical and vision systems. Our clients accepted 186.4 million items, and 98.7% of what we delivered was accepted the first time.4 We name a client only with its permission.
| Image and video | 58% |
|---|---|
| Text and language | 22% |
| Sensitive-content review | 12% |
| Model evaluation | 8% |
Rungwell at Home
940 labellers worked from home at the end of 2025, on Rungwell laptops with data paid by us. Rungwell at Home began as an emergency in 2020 and has become one of the best things we do: it reaches parents of young children, people with disabilities and people in towns too far from Kilimani or Cubao to commute.
Remote work is only possible because clients trust it, and Focus Assurance is a large part of why they do. It helps labellers notice when their attention drifts, suggests breaks, and gives clients evidence that work done at home meets the same standard as work done in our centres. In 2025, remote and on-site quality scores were within 0.4 points of each other, and labellers accepted 62% of the break suggestions Focus Assurance made.9
| Kenya | Share | Philippines | Share |
|---|---|---|---|
| Nairobi County | 41% | Metro Manila | 38% |
| Kiambu | 22% | Rizal | 19% |
| Machakos | 11% | Bulacan | 16% |
| Kajiado | 9% | Cavite | 14% |
| Nakuru | 6% | Laguna | 8% |
| Elsewhere | 11% | Elsewhere | 5% |
The money
Rungwell earned CA$38.6 million in 2025, 96% of it from clients and 4% from foundation grants for training. As a not-for-profit we distribute nothing: the surplus goes back into the organization.10
| CA$ m | Share | |
|---|---|---|
| Labellers’ pay, benefits and statutory contributions | 23.5 | 61% |
| Centres and facilities | 5.4 | 14% |
| Training and wellbeing | 2.7 | 7% |
| Technology, including Rungwell Tasks | 2.3 | 6% |
| Administration and governance | 3.1 | 8% |
| Surplus, reinvested | 1.6 | 4% |
| Total | 38.6 | 100% |
Where we fell short, and what we will do
- Retention. Of the people who joined in 2024, 64% were still with us a year later, against our target of 70%. Most who leave go in the first three months. In 2026 we will extend paid training by a week and give every new labeller a named mentor from the cohort before theirs.
- What happens after Rungwell. Fewer than four in ten leavers answer our survey. In 2026 we will offer a reference letter and a skills record to everyone who leaves, and ask for their news when we send it.
- Wellbeing on a busy shift. Labellers on sensitive work value the program and find it hard to use when the queue is long. We will commission an independent review of the wellbeing program in 2026 and publish what it finds.
- Pay, explained. Labellers have asked us to explain more clearly how earnings are worked out. In our 2026 report we will publish the range of task earnings by type of work.
Notes
- Everyone employed by Rungwell Kenya Limited or Rungwell Philippines, Inc. in a delivery role between 14 March 2016 and 31 December 2025, counted once however many contracts they held.
- Median gross monthly earnings (base pay for scheduled shifts plus task earnings) of labellers who worked a full month, divided by the statutory minimum wage that applies to the work in Nairobi, and in Metro Manila for a 26-day month. A full month is 20 or more scheduled shifts with an average quality score of 95% or more (on site) or an average Focus Assurance score of 80 or more (Rungwell at Home). In a typical month in 2025, 58% of labellers worked a full month. Task rates vary by task.
- Promotions to senior labeller, quality reviewer, trainer, team lead or centre management since 2016. A person promoted twice is counted twice.
- Items accepted by the client at first delivery, as a share of items delivered, across all projects in 2025. Items returned by a client and corrected are counted as not accepted.
- “No formal work” means no formal employment in the 12 months before applying, as stated on the application form. Disability is self-identified in a voluntary question asked at hiring and again each year.
- Senior labellers and quality reviewers receive a higher base for each scheduled shift. Quality reviewers’ task earnings are for review work. Task rates vary by task and are shown to labellers in Rungwell Tasks before a task is accepted.
- The survey is sent 12 months after a person leaves. Of the 811 people who left Rungwell in 2024, 309 replied (38%). People who are in work may be more likely to reply than people who are not; we have not weighted the results.
- Wellbeing time is 2 hours a month for labellers on S1 work, 4 hours on S2 and 8 hours on S3, taken in the shift and credited at the centre’s standard task rate. Uptake is hours used as a share of hours available to S2 and S3 labellers, averaged over the year.
- Focus Assurance scores are calculated on the labeller’s laptop and range from 0 to 100; see our worker privacy notice. Items completed while a session’s focus score is below 70 are sent for a second review before delivery and are credited when released.
- Unaudited summary of the consolidated statement of operations. The audited consolidated financial statements for the year ended 31 December 2025 are on our financial statements page.
- Promotions by year count the year in which the new role began.
Questions about this report: press@rungwell.org. How we collect and check these figures: How we measure.